On average, only a little, and not enough to eat back the workout. After a single session, people ate about the same as on a rest day2. Over weeks of training, most trials found no change in daily food intake, though hunger before breakfast can rise slightly4.
"On average" is doing work in that answer. Some people do eat more after they start training. Below: what was measured, what was found, and why individual experience can differ from the averages.
What was studied
Researchers ask two separate questions.
One workout. People exercise, or rest, on separate days. Then they are offered a meal to eat freely (ad libitum, in the papers), and the leftovers are weighed. Hunger is rated on a 100 mm line from "not at all hungry" to "extremely hungry".
Weeks of training. People follow an exercise programme for weeks or months, and their daily food intake is compared before and after, or with a group that did not exercise. Daily intake is usually self-reported with food diaries.
A few more terms:
| Term | Plain meaning |
|---|---|
| Compensation | Eating extra food that replaces the energy a workout used |
| Meta-analysis | One analysis that pools many trials |
| Systematic review | A search for every study on a question, read by set rules |
| Fasting hunger | How hungry people rate themselves before breakfast |
| Effect size | A difference expressed on a standard scale, so different studies can be pooled |
In plain terms: the question is whether exercise leads to eating more, measured at the next meal and over weeks.
What the studies found
After one workout: almost no change. A 2013 meta-analysis pooled 51 trials from 29 studies2. Sessions lasted 30 to 120 minutes, and a free meal was offered up to two hours later. Food eaten afterwards changed trivially (effect size 0.14, 95% confidence interval −0.005 to 0.29). A confidence interval is the range the true value most likely sits in; this one just touches zero.
Counting the energy the workout used, the picture shifts. Intake minus exercise cost was much lower on exercise days (effect size −1.35)2. In the authors' words, people "tend not to compensate" in the hours right after exercise.
Exercise and eating: pooled effect sizes · effect size
Weeks of training, healthy adults: mostly no change. A 2014 systematic review of 99 articles in healthy, non-athlete adults aged 18 to 64 did not pool the numbers, because the studies were too varied3. It counted them instead. Of 24 randomised trials of exercise training, 18 found no significant change in daily food intake. Five found decreases, of about 200 to 500 kcal a day. One found an increase, in one subgroup only.
Randomised training trials by effect on daily food intake · trials
Weeks of training, adults with overweight or obesity: hunger up a little, intake flat. A 2021 meta-analysis gathered 48 articles, with training lasting a median of 12 weeks4. Across 25 study arms, daily intake after training did not differ from non-exercising control groups: a mean difference of −13 kcal a day (95% CI −83 to 58). In the five arms from studies rated fair or good quality, exercisers ate about 102 kcal a day more (95% CI 1 to 203), a difference the authors rated small to negligible4.
In the smaller set of better-quality studies, exercisers ate about 102 kcal a day more than controls (95% CI 1 to 203)4. The authors rated that difference as negligible in size.
Daily intake after training, exercisers minus controls · kcal/day
Hunger before breakfast rose by about 8 mm on the 100 mm line after training (95% CI 4 to 11), a small effect4. That comparison was before vs after within the exercise groups, not against controls. Fullness before breakfast did not change.
In plain terms: exercise may nudge morning hunger up a little, but on average people did not eat much more.
Why some people feel hungrier
Averages hide the spread. In a 2008 study, 35 sedentary adults with overweight or obesity did supervised exercise five times a week for 12 weeks1. The researchers split them by how their results compared with prediction. In one group, daily intake rose by about 268 kcal and daily hunger by about 7 mm. In the other, intake fell by about 130 kcal and hunger did not change.
The spread within each group was wide too: the standard deviations were larger than the averages1. So "exercise does not make people eat more" and "I eat more when I train" can both be true.
Intensity and timing may matter. Most single-session studies used moderate exercise and a meal soon after2. The 2021 review's authors describe a model in which training raises hunger a little while also strengthening fullness after meals4. Their data fit it, but it is not settled.
Hunger and eating are not the same. Feeling hungrier before breakfast did not translate into higher daily intake in the pooled trials4. How hungry you feel and how much you eat are linked but separate.
In plain terms: the average person barely changes, but individuals can move a few hundred calories either way.
What this means for tracking
These nutrition and training trials describe averages, not a plan. The workout's energy cost is often smaller than people expect, which is why exercise burns fewer calories than the tracker says. Everyday movement outside training matters too: see what NEAT is.
Because individual responses vary, the most useful data is your own. A food log alongside a training log shows whether intake drifted up after you started, rather than guessing from averages. For the energy-balance frame behind all of this, see calories in, calories out.
The studies at a glance
| Study | Who was studied | Key result | Year |
|---|---|---|---|
| King et al., *Int J Obes*1 | 35 sedentary adults with overweight or obesity; 12 weeks, 5 sessions a week (abstract) | Wide individual spread: one group's intake rose about 268 kcal/day, the other's fell about 130 | 2008 |
| Schubert et al., *Appetite*2 | 51 trials in 29 studies; single sessions of 30–120 min (abstract) | Food eaten afterwards: trivial change (effect size 0.14); no eating back of the session | 2013 |
| Donnelly et al., *PLoS One*3 | Healthy adults aged 18–64; 99 articles, 24 randomised trials | 18 of 24 trials: no change in daily intake; 5 less; 1 more | 2014 |
| Beaulieu et al., *Obes Rev*4 | Adults with overweight or obesity; 48 articles, median 12 weeks | Daily intake vs controls: −13 kcal/day; fasting hunger +8 mm (small) | 2021 |
| Launbo et al., *EBioMedicine*5 | 58 adults with overweight or obesity, with and without type 2 diabetes; crossover | Ate about 86 kcal less after a 45-min interval session than after rest; no catch-up over 24 h | 2026 |
Limits worth knowing
- Self-reported intake. Most training studies relied on self-reported daily intake, which the 2021 authors list as a limit4.
- Few trials built for this question. The 2014 review names a lack of adequately powered trials that also measured intake well3.
- Low study quality. In the 2021 review, 39 of 48 articles (81%) were rated poor quality4.
- Different people in different reviews. The 2014 review covered healthy adults; the 2021 review adults with overweight or obesity. Results may not carry across.
- Abstract only for two papers. We read the 2013 meta-analysis and the 2008 study as abstracts; the others in full.
- Funding. The 2026 trial was funded by a pharmaceutical company5.
Latest research
The newest trial we opened is from 20265. In Copenhagen, 58 adults with overweight or obesity, half of them with type 2 diabetes, came in for four lab visits. On two they cycled for 45 minutes, including four four-minute hard intervals; on two they rested. Visits were in the morning or late afternoon.
After exercise, they ate 361 kJ less at a free meal than after rest (95% CI 202 to 520 kJ), which is about 86 kcal5. They did not make it up over the next 24 hours, and appetite ratings were lower after exercise. Time of day made little difference overall. It is one session per condition, but it fits the pattern: a hard workout did not lead to eating more that day.
The short version
- After one workout, people ate about the same as on a rest day, so they did not eat the session back.
- Over weeks of training, most trials found no change in daily food intake.
- Hunger before breakfast can rise a little with training, by about 8 mm on a 100 mm scale.
- Individuals vary widely: some eat a few hundred calories more, others less.
Educational overview only — not medical advice.
